24 ILUNOIS BIOLOGICAL MONOGRAPHS [24 



SO that the cranium does not end in the same place in all of these groups. 

 The cranial-vertebral complex of the Amniotes includes three vertebrae and 

 the hypoglossus nerve of this group is the result of the fusion of the three nerves 

 corresponding to these three vertebrae. He says, that in a general way the 

 auximetameric neocranium of the teleost agrees with the Amniotic condition. 



The nerve formula for the occipital region of the Siluroids is b-0-4. Accoi d- 

 ing to this view, the first post-vagal nerve of Amiurus is an 'occipito-spinale' 

 nerve and corresponds to the second free spinal nerve of the Selachians, which 

 the Arabic numeral represents. Thus, there is one segment missing between 

 the paleocranium of Amiurus and the auximetameric neocranium, and another 

 between this and the first free spinal nerve. This method of reasoning is based 

 on Sagemehl's hypothesis that in the Characinidae and the Siluridae, where the 

 Weberian apparatus is developed, the claustrum represents the neurapophysis 

 of a rudimentary vertebra, and that the nerve which originally came in be- 

 tween the claustrum and the scaphium is lost, together with the muscle seg- 

 ment. I do not regard the claustrum as a rudimentary vertebra, but as an 

 intercalated cartilage, developed in connexion with the specialized Weberian 

 ossicles. The scaphium may be the first true neurapophysis and a modified 

 representative of this part of the first vertebra. I hope to get further evidence 

 later for the exact somitic relations of these parts in younger larva than have 

 yet been accessible. For the present I accept the hypothesis suggested by 

 Kingsley (1910) concerning the relations of the occipital region in vertebrates, 

 where he says: "In the vertebrates there is a continuous addition of new so- 

 mites at the posterior end of the body as in the arthropods and annelids, im- 

 plying the existence of the equivalent teloblasts at the posterior end. The 

 assumption of budding zones at other points will explain other features noted. 

 Such a zone in the occipital region will allow us to explain the difference in the 

 number of cranial nerves in the Mammals and the Ichthyopsida and yet allow 

 us to accept the homology of the occipital bone throughout the series. The 

 additional nerves are thus to be regarded not as transferred from the neck, 

 but as new or intercalated structures. " 



Reasoning on this basis, the first post-vagal nerve of Amiurus is a new, inter- 

 calated formation, possibly associated with the second somite. What part the 

 first somite plays in the development of the occipital region I have not been 

 able to ascertain, if we assume that the first dorsal muscle segment is that of 

 the first metotic somite. In the larval Salmo, MissWillcox found that there 

 were five segments between the posterior end of the otic capsule and the first 

 neurapophysis. Two of these disappear very early in development and have 

 no trace of nerve connexions. The third has a rudimentary nerve which atro- 

 phies early. The next two nerves innervate the next two somites and issue 

 from the cranium between the parachordals and the neurapophysis of the 

 first vertebra. These are later enclosed in bone and leave the exoccipital 

 through the same foramen as the hypoglossus. Fiirbringer's formula for this 



